Nico Huber has posted comments on this change. ( https://review.coreboot.org/c/coreboot/+/32407 )
Change subject: soc/amd/picasso: Create picasso as a copy of stoneyridge ......................................................................
Patch Set 1:
(2 comments)
https://review.coreboot.org/#/c/32407/1//COMMIT_MSG Commit Message:
https://review.coreboot.org/#/c/32407/1//COMMIT_MSG@12 PS1, Line 12: it's the most transparent
The biggest argument is, that the present code has been reviewed,
This is true.
and does not have to be reviewed again.
This is very subjective. As somebody who takes what the code does into account and not only how it looks like, I state the exact opposite: It has to be reviewed again (for the new silicon).
I have seen it very often now that after the first train of copy- first patches, the development seems to turn into a bug fixing because nobody keeps track any more which part of the code has already been reviewed/adapted for the new silicon. So if something doesn't work, I assume a developer will start digging through the code and look for things that might have to be adapted. Is that documented somewhere that he did so? I doubt it. So the next one with an issue has to go through everything again, just because the information is lost what parts are already re-reviewed. In the long run all this information is completely lost, even though we actually chose Git to preserve such knowledge.
That's just the Git-history view on the matter. I could go on about how the code bases tend to deviate and what that means for maintenance...
https://review.coreboot.org/#/c/32407/1/src/soc/amd/picasso/smbus.c File src/soc/amd/picasso/smbus.c:
PS1: Nice code, wow you added 237 lines to coreboot, thanks! Maybe you missed that we already have half a dozen copies of it?